Why I'm A Molinist

Non-Molinism as the "but I did eat breakfast" answer.

In his famous treatise Logic and Theism, J.H. Sobel titled his fifth chapter “God Knows (Go Figure).”

But what exactly does God know? Atheists, will of course, answer nothing, but for those of us who believe in an omniscient God, what precisely does he know? Everyone agrees that he knows, for instance, all of the mathematical facts and the number of apples on earth. But what does God know in relation to our freely willed choices? There are four main views on this:

1. Theological determinism: our actions are predetermined by God. He picks what we do. So he knows them because he directly determines them. This view is adopted by Calvinists—go figure!

2. Open theism: on this view, God doesn’t even know the complete future. If we have free will, on such a view, God doesn’t know in advance what we’ll do. So as we freely make choices, God learns that we freely made the choices that we in fact make. He didn’t know in advance, for instance, that I’d choose to write this sentence rather than a different one, or that Gandhi would freely choose to do the right thing and Hitler the wrong thing.

3. Simple foreknowledge: On this view, God knows the future. At the dawn of time (or perhaps timelessly, if God is outside of time), he knew everything that would ever happen.

4. Molinism: On this view, not only does God know what we will do, he knows what we would do under different conditions. So he knows, for instance, what I would do if you plopped me in ancient Egypt or what I’d do if a random stranger went up to me and slapped me. God knows what we in the biz call counterfactuals of freedom—those are what agents would freely choose in counterfactual circumstances. Before making the world, God knew what every possible agent would do in every state he could have made it in.

I lean towards Molinism. While there are lots of objections to it, many of them quite clever and sophisticated, to me they strike me as interesting puzzles rather than decisive objections. I think there’s a core intuition behind Molinism that strikes me as pretty decisive. I don’t intend this article to be exhaustive (or exhausting) but just to provide a brief explanation of why Molinism has always struck me as sensible.

Theological determinism, according to which God predetermined our actions, is very hard to believe. Why would God predetermine Hitler to mass murder Jews or Genghis Khan to go on mass killing and raping sprees? If there are evil actions—which, of course, theists should accept—why would God predetermine people to take those actions? Furthermore, it seems like if something else predetermined your actions, you’re not really free. You’re a dancing marionette on a string, not genuinely in control of your acitons. Finally, I think once the theist accepts that God is free—which they should, to account for his choice to create whichever particular world he did, rather than a different one—accepting that we are free isn’t much of an extra cost.

For this reason, I think the serious competitors to Molinism are mostly open theism—according to which God doesn’t know free choices in advance—and simple foreknowledge, according to which God exhaustively knows the futures but no the counterfactuals of freedom (he knows what I will do but not what I would do if deposited in Rome). Simple foreknowledge is the most formidable competitor. But to see what’s wrong with it, let’s first see what’s wrong with open theism.

Open theism seems defective because it arbitrarily constrains divine omniscience. It’s weird for God to know everything except a few facts—it would be like God knowing every fact except the facts about calculus, which, while making God a lot more personally relatable, would be bizarre and implausible. So the open theist will probably have to deny that there are facts about the future. But it really seems like there are facts about what will happen in the future.

For example, today I ate breakfast. It seems like for that very reason, yesterday it was a fact that tomorrow (relative to yesterday) I’d eat breakfast. Similarly, tomorrow I’m getting on a plane to the United Kingdom—that sure seems like a fact, even though it will ultimately depend on the free choice I make tomorrow. But if there are facts about the future, then it seems like God could know those facts. Facts are just things that are the case—accurate descriptions of reality—but there’s an accurate description of the reality of what I’ll do in the future.

So the basic argument against simple foreknowledge seems to be straightforward—to paraphrase:

1. There are facts about agents will freely choose (E.g. it’s a fact that I’ll freely choose to get on a plane tomorrow).

2. If there are facts about what agents will freely choose, God would know them (he’s omniscient!)

3. So God would know what agents will freely choose.

I also don’t think there’s a good argument for open theism. Generally, people argue for open theism by claiming that if God knows what we’ll do in advance then we’re not free to do choose because we can’t do other than what God knows we’ll do. A necessary component of free will is being able to do otherwise. But I don’t think this strips us of free will—God knowing I’ll board my plane tomorrow doesn’t force me to board it. Instead, God knows I’ll board it because I will actually board it. Something can’t eliminate my freedom to choose if my freedom to choose explains why the thing happens! The fact that in 3,000 BCE it was true that I’d board my plane doesn’t mean I can’t do otherwise—it just means that if I were to do otherwise, the fact in 3,000 BCE would have been different. God predetermining our actions would strip us of free will, but that’s because then his actions would explain what we do, not the other way around. This point is defended well by Zagzebski (in class, at one point, rather embarrassingly, I hadn’t done the reading which was by Zagzebski, but had had similar thoughts to Zagzebski, and so then I, thinking I was being original, during the class discussion just said the same things Zagzebski said. Very embarassing—many eggs were thrown and people said of me “who is that fool who simply regurgitated the reading as if it was an original point?”)

But I think this same basic kind of argument can be made against simple foreknowledge—the view that God knows the future but not counterfactuals of freedom.

It sure seems like there are facts about what people would do in different circumstances. For example, though I don’t plan on poking my plane neighbor in the eye while on the flight, it seems like there’s a fact of the matter about what would happen if I did that. It’s a conceivable scenario—something would happen if I did it—so it seems like there should be a fact of the matter about what would happen if I did it. But if there’s a fact of the matter, then God could know it.

We make counterfactual statements like this all the time. We say things like “if I’d asked her out, she’d have said yes.” It seems odd to think that in cases like that, there really is no fact of the matter about what a person would have said if asked out.

This point becomes all the more clear when we take into account God’s choices depending on the future. It seems perfectly reasonable for God to peek into the future, see that, say, Hitler will do a bunch of bads things (I double checked this with a historian and he said it was right), and then decide not to create Hitler. But then God would no longer know what Hitler would do, so he’d no longer be able to peer into the future. The simple foreknowledge view becomes very hard to make sense of when God’s choices will depend on his knowledge of the future—if his decisions change based on learning something about the future, then that would change the future, meaning that he couldn’t have originally known the future to begin with, as what he knew was no longer the future.

The only way out of this that I can imagine—cleverly suggested to be in a conversation with the wonderful Philip Swenson, one of my favorite living philosophers—is to suggest that God can only peek into the future after he’s fully decided what he will do. So, for example, he can only see what Hitler will do after resolved to create Hitler.

But this seems to open up its own can of worms in two ways. First, it seems like an odd constraint on God’s power. If God learns that Hitler will do bad things, what could possibly stop him from choosing to zap Hitler out of existence during the first moment of Hitler’s creation?

Second—more concerningly—the core problem for non-Molinist views seems even more pronounced. Suppose that God has two options—A and B—of what to create. He has not yet peeked into the future to see how history will play out. Only after agreeing to make, say, A can he know what will happen if he produces A.

But this seems super weird! Something will happen if he creates A. Suppose—this will assume God making his decisions in time, but similar points apply to a timeless God—God is reflecting at 10:00 am about which world to create. Assume that he’ll make the choice to either create or not at 10:15. On this picture, God doesn’t know what will happen if he creates either A or B at 10:00.

But something will happen. The world will play out some way if he creates A, and a different way if he creates B. Why couldn’t he know how it would play out? This seems, like the open theist view, to arbitrarily restrict his knowledge—to make it so that he doesn’t know certain things, despite his essential omniscience. Thus, we can argue in such a case:

1. If God creates A, something specific will happen.

2. If something will happen if God creates A, God could know what will happen.

3. Therefore, God could know what will happen if he creates A.

This argument strike me as quite strong. While there are some interesting and tricky puzzles for the Molinist—as well as for other views—to my mind, they are much less powerful than the core Molinist intuition, that there are facts about counterfactuals of freedom and that God could know those facts.

Expect posting to be slow—I just arrived in the UK for the start of the school year (though I wrote the post before getting here, hence the passages about having a future plane ride).